Even Andreas Jansson Nilsen Alexander Nicolai Askevold Storkaas
BI Norwegian Business School - Preliminary Thesis -
Talking the Talk: An Empirical Investigation into the Economic Effects of Strategy Disclosure
Supervisor:
Amir Sasson
Hand-in date:
13.01.2017
Campus:
BI Oslo
Examination code and name:
GRA 19501 Preliminary Thesis
Programme:
Master of Science in Business, Major in Strategy
This thesis is a part of the MSc programme at BI Norwegian Business School. The school takes no responsibility for the methods used, results found and conclusions drawn.
EXECUTIVE SUMMARY... II
INTRODUCTION AND RESEARCH QUESTION: ... 1
LITERATURE REVIEW: ... 2
STRATEGY AND CLASSIFICATION ... 2
STRATEGY AND DISCLOSURE ... 3
IMPLICATIONS FOR OUR RESEARCH: ... 5
CRITIQUE: ... 6
HYPOTHESIS: ... 7
RESEARCH DESIGN AND METHODOLOGY ... 7
BRIEF INTRODUCTION TO CONTENT ANALYSIS. ... 7
CONTENT AND CLASSIFICATION ... 8
RELIABILITY ASSESSMENT ... 8
TOTAL ASSESSMENT OF RELIABILITY: ... 9
RELIABILITY DATA:... 9
VALIDITY ... 11
CREATING AND TESTING A CODING SCHEME –WEBER PROTOCOL ... 12
MEASUREMENT MODELS ... 12
CHALLENGES RELATED TO THE SCHEME ... 13
SAMPLE SELECTION ... 13
PROJECT ORGANIZATION AND TIMELINE: ... 14
REFERENCES: ... 16
APPENDIX ... 20
Executive Summary
This paper aims to further the research on the economic effects of strategy disclosure in annual reports. Using content analysis, the authors will examine the communication in corporate annual reports for listed firms on Oslo Stock
Exchange, and construct a scheme to rate the firms on their quantity and quality of strategy communication. Despite the implied importance of disclosure in
academic research, strategy disclosure remains a highly underexplored area of research. This analysis will attempt to fill parts of that gap in the literature.
As corporate annual reports are recognized as a prime medium of information for investors and other stakeholders, research into the voluntary disclosure of non- financial elements of the firm has increased in importance in recent years. While most disclosure research has focused on general content or corporate social responsibility, however, little attention has been devoted to the role of strategy.
Building on this, our paper will explore the effects of increased strategy disclosure on different dimensions of firm value.
In order to examine this effect, we will construct a comprehensive scheme to rate firms on different aspects of strategy disclosure through careful content analysis of their annual reports. Using these ratings, we can follow a quantitative approach to explore their effects on firm value, where data for listed firms are widely available. As the quality of our research relies substantially on the merit of our scheme, much time will be devoted to its construction and subsequent tests for validity and reliability.
In conducting this research, the authors hope to contribute to an important, yet partially neglected, area of research.
Introduction and Research Question:
This paper aims to further the research into strategy disclosure in corporate annual reports, and examine the relationship between disclosure of different strategic dimensions and firm value. Through content analysis of firms’ annual reports, we will construct a scheme to rate and classify firms on the quantity and quality of their strategy disclosure, which will provide insights into the communicative importance of annual reports. In order to examine this effect, we will treat strategy as the desired future state of the firm, and its path to get there.
Today, corporate annual reports are considered an important informative tool for investors and other stakeholders, providing factual insights and reducing
information asymmetries. Further, recent legal initiatives have increased the demands facing firms regarding the information disclosed, while the accessibility of annual reports have extended substantially with the technological advances of the last decades. This has led to an important role for disclosure research in the academic literature, as insights could potentially have important implications.
Despite this, previous academic foci have centered around general disclosure effects and corporate social responsibility on different firm characteristics, with strategic disclosure representing only a limited part of this increasing literature.
Considering the academically implied importance of strategic disclosure in
corporate communication, the hitherto underwhelming amount of research into the field opens possibilities for exploration. While annual reports contain satisfactory content on the financial situation of firms due to legal requirements, corporations do not face the same demands regarding strategic discourse. Instead, insufficient time and corporate resources are allocated to the communication of strategic initiatives in annual reports, representing a principally disregarded area of
research. The authors of this paper will, humbly, aim to fill parts of this gap in the academic sphere through our analysis, attempting to shed light on an insufficiently researched field. This leads us to the research question guiding our study:
“How does disclosure of strategy in annual reports affect firm value?”
Specifically, we aim to construct a scheme to represent both the quantity and quality of strategy disclosure in annual reports for firms on Oslo Stock Exchange,
attempting to rate them on different aspects related to their communication of strategy. In doing so, these ratings will allow us to inquire the effects of strategic disclosure on different dimensions of firm value, thus investigating the role of strategy communication to stakeholders and its economic effects on the firm.
Literature Review:
Strategy and Classification
Strategy can be seen as the ongoing stream of decisions that guide the
organization’s continuous alignment with its environment, internal policies, and procedures. The multidimensional nature of strategy, and its situational
contingency, managed for a long time to elude any precise operationalization and classification of the concept (Hambrick, 1983). Starting with Miles and Snow´s (1978) Typology, which enabled the discrete classification of business level strategies across industries, a stream of empirical research followed. This initially focused on the validity of the concept, but soon developed into focusing on establishing a link between strategy and firm performance (See Dvir et al., 1993;
Hambrick, 1983; Shortell & Zajac, 1990).
Miles and Snow’s (1978) Typology hypothesized that organizations enact their environments, and become dominated by the adaptive decision patterns they employ (James & Hatten, 1995, p. 161). On the basis of these patterns, they formed discrete categories that, according to Miles and Snow (1978), illustrate four strategic archetypes; Defenders, Prospectors, Analyzers, and Reactors. A major obstacle to research at the business level is that industries have their own unique environmental attributes and strategic options (Hambrick, 1983). The link between strategy and performance was at this time in high fashion in academia, but most research was limited to single industries, such as Datta’s (1979) study of the TV industry, or Hatten et al.’s (1978) study of the US brewing industry. Miles and Snow’s (1978) Typology, on the other hand, allowed interindustry research, thus reducing the potential issue of external validity.
The Typology created a great deal of interest and debate, and the construct validity of the self-typing paragraph method often used to classify the sample set has been well accepted (James & Hatten, 1995). However, Hambrick (1983) noted
that “… typologies represent a theorist´s attempt to make sense out of non-
quantified observations. They may have the advantage of being “poetic”,… that is ring true, often sounding very plausible. However, since they are the product of rather personal insight, they may not serve well for descriptive purposes but have limited explanatory or predictive power” (Hambrick, 1983, p. 28). This is also the main objection of DeSarbo et al.’s (2005) recent revisit of the topic, using a modified constrained multi-objective classification method (NORMCLUS), developed by DeSarbo and Grisaffe (1998). An inductive, as opposed to the deductive approach of Miles and Snow (1978), is taken. Comparing the
classification obtained from the clustering procedure, DeSarbo et al. (2005) find a somewhat different classification. They do not, however, negate those promoted by Miles and Snow. DeSarbo et al. (2005) describe their finding as a second order derivative of the pure and conceptually distinct Miles and Snow (1978) Typology, which in this case is considered a first order “primitive”. The notion that the data shows a more nuanced picture is hardly surprising, and is likely the reason why typologies such as Miles & Snow (1978) or Porter´s (1980) generic strategies is seldom used in quantitative research today (For an application of Porter´s generic strategies, see; Dess and Davis (1984).
The general idea presented by Miles and Snow’s (1978) Typology, however, provided the foundation of further attempts to quantify and classify firm strategy in academia, and lays as a cornerstone for disclosure research.
Strategy and Disclosure
During the 1970s and -80s, Professor Edward H. Bowman of Sloan School of Management wrote a series of articles that examined the content of firms’ Annual Reports (Bowman, 1976, 1978, 1984), arguing that the scrutiny of these could provide insights into the effectiveness of a company’s strategy. Through careful content analysis, Bowman deducted behavioral differences between well-
performing firms and their underachieving equivalents, indicating the informative value of annual reports as an important tool for investors and stakeholders alike (Kohut & Segars, 1992). In his articles, Bowman employed a line-by-line comparison of report content for different firms, building on the idea of strategy as something quantifiable and providing an early venture into disclosure theory.
Annual Reports are recognized as a prime tool for investor decision making (Benartzi & Thaler, 1993) and companies can use it strategically as a
communication medium for different stakeholders (Stanton & Stanton, 2002). It allows a company to proactively paint an external picture of its own existence, with Hines (1988) arguing that in “…communicating reality, you construct
reality,” (Hines, 1988, p. 257). While the ostensible content may conceivably hold little resemblance to the de facto state of a given firm, annual reports are regarded as a powerful source of information regarding company matters (Diamond &
Verrecchia, 1991).
This significant role has led to research into the disclosure level of annual reports, examining the relationship between the quality and quantity of content and different firm characteristics (Ullmann, 1985). Even as companies must conform to certain regulatory constraints and demands regarding content, some go beyond the legal imperative. This voluntary disclosure – defined as that in excess of the required – has thus become the subject of intensive research, building on the premise of its informative value (Meek et al., 1995). Despite Bowman’s early research into the effects of disclosure of strategic actions, subsequent literature has mainly focused on the effects of disclosure level in general or social
disclosure on firm performance and characteristics (Santema & Van de Rijt, 2001;
Stanton & Stanton, 2002). The former includes, but is not limited to, research on the effect of increased disclosure quality on the equity cost of capital (Botosan, 1997; Hail, 2002), disclosure quality on stock returns and liquidity (Diamond &
Verrecchia, 1991; Healy et al., 1999), disclosure comprehensiveness and different firm characteristics (Terence E Cooke, 1989; 1992; Wallace et al., 1994), as well as the choice of accounting standards on bid-ask spreads and trading volume (Leuz & Verrecchia, 2000). The latter has mainly focused on social disclosure as determinants of firm’s Return on Equity through self-constructed scores (Abbott
& Monsen, 1979) or as percentage of annual reports (Bowman, 1978), and its effect on a firm’s social performance – an organization’s responses to social demands (Strand, 1983). This latter relationship has been hypothesized to subsequently effect the economic performance of firms, albeit with mixed academic evidence (Patten, 1992; Ullmann, 1985).
While no consensus regarding voluntary disclosure has yet been reached, evidence indicates that the presupposition that its increase has positive effects is well
grounded, as disclosure is assumed to reduce the informational uncertainty and asymmetry facing investors (Healy & Palepu, 2001; Ullmann, 1985). As most academic focus is guided to disclosures regarding accounting and corporate social responsibility, however, little attention has been given to the Executive narrative part and subsequent strategic revelations found in corporate annual reports (Abrahamson & Amir, 1996; Santema et al., 2005) despite Bartlett and Chandler (1997) and Barry and Elmes (1997) underlining its integral role for shareholders and potential investors.
Following the early research by Bowman on the role of strategy in annual reports, Higgins and Bannister (1992) argued that strategic credibility, partly achieved through revelations in annual reports, affected a company’s share price,
encouraging further research into corporate communication on strategy. Barron et al. (1999) found that higher Management Analysis & Discussion (MD&A) ratings from annual reports were positively correlated with less dispersion and error in earnings forecast by analysts, with regression estimates showing MD&A quality as a powerful tool for predicting future earnings (Barron et al., 1999). Further, research into strategy disclosure in annual reports of Dutch firms (Santema & Van de Rijt, 2001) and, by extension, firms across Europe (Santema et al., 2005) found that firms in general disclose relatively little regarding strategy, as opposed to financial statements, while also showing that the amount of disclosure differ across countries (Santema et al., 2005). Although academic insights have suggested positive effects of increasing levels of strategy disclosure, research in the area remain scarce.
Implications for Our Research:
Appreciation for the conceptual developments of the aforementioned Typologies and the idea of strategy as something quantifiable is vital to the development of our own classification, as the methods which we base our own research on can be found to be similar in many regards. We find, however, the concept of using a discrete classification to be inadequate, especially for rating/barometer purposes – the implicit assumption of constant and equal dissimilarities between categories is
perhaps the strongest motivating factor for creating a continuous rather than discrete measure. The choice of using a continuous classification also has implications for measuring methods, and invalidates, in our opinion, the self- typing classification method. The primary concern rises from the issue with different perceptions of relative distance between score levels. This is also the reason why output from the similar Likert-scale is normally treated as ordinal data, rather than interval data (D. Bertram, 2007) – a prerequisite for the construction of a continuous scale.
Critique:
One concern with regards to the classification and quantification of strategic dimensions relates to their ambiguity, and the subjective interpretation of the theorists – as proposed by Hambrick (1983), any such scheme risks suffering from limited predictive power and biasedness. The idea of reducing multidimensional variability between firms introduces important limitations, and can potentially result in overgeneralized classifications.
A main critique related to voluntary disclosures regards the nature of the
disclosure itself. Non-financial disclosure in e.g. annual reports are decided on ex post, where the firm decides what information to disclose only after it is observed (Leuz & Verrecchia, 2000). Further, voluntary disclosures on future strategic initiatives do not necessarily represent firm commitments, as they can be reviewed ad hoc or reversed over time, potentially reducing the validity of the statements given (Diamond & Verrecchia, 1991). Another issue is the accuracy and
verifiability of the information given in the reports, as executives may have different incentives than other stakeholders (Gu & Li, 2007) – it might not be in their best interest to provide an entirely genuine representation of reality.
Nevertheless, Gu and Li (2007) found that voluntary disclosure statements on innovation strategy had significant positive effects on firms’ stock market prices, indicating that investors generally perceived the strategic disclosures of firms to be credible.
In spite of these important and well-founded arguments, we argue that a scheme as proposed in this paper will be able to capture important elements of strategy
communication, providing insights into its effect on firm value and encourage further research in this valuable field.
Hypothesis:
Following seminal research, this paper will use a scheme to classify and rate firms based on their communication of strategy to different stakeholders. As previous research has shown potential benefits from corporate disclosure in annual reports, it would seem natural to assume the same regarding strategic disclosure.
Additionally, annual reports are recognized as a prime medium for obtaining information related to corporate strategic initiatives, thereby allowing investors to make an informed prediction as to the future direction of the firm. Thus,
considering the potentially reduced information asymmetries and improved earnings forecast associated with increased disclosure, we would expect to find a positive relationship between strategy disclosure and firm value. We provide the following hypothesis for our analysis:
Hypothesis: A higher strategy disclosure rating will, ceteris paribus, lead to higher firm value.
Research Design and Methodology
Brief introduction to content analysis.
As defined by Neuendorf (2002, p. 1), content analysis is; “…the systematic, objective, quantitative analysis of message characteristics,”. Further, Habermas (1987, p. 80) states that “…we need to note that communicative action rest at the very base of the lifeworld, and one very important way of coming to grips with that world is to study the content of what people say and write in the course of their everyday life,”. On the other hand, where methods borrowed from the natural sciences have been applied, social researchers prevent themselves from addressing what matters most in everyday social life; human communication, commitments people make to each other and to the conception of society they aspire to, what they know, and why they act (Krippendorff, 2004, p. 11). Certainly, content analysis is not the only research method that seeks to capture what is mediated
between people, texts, information, symbols so forth, but it has developed over the years into one of the strongest tools for interpreting communication.
Content and classification
To make valid inferences from text, it is important that the classification
procedure is reliable in the sense of being consistent; different individuals should code the same text in the same way (Weber, 1990, p. 12). For our research, we will be two coders, so in order to limit any bias that might arise from differences between us routines to control intercoder reliability is important. Classification by multiple human coders permits the quantitative assessment of achieved reliability (Weber, 1990, p. 14), and implementation of these routines are essential to the quality of our research. Much harder to assess will be the validity of variables based on content classification. The following section will discuss these different themes.
Reliability assessment
Krippendorff (2004, pp. 211-221) describes three highly pertinent issues related to the reliability of content analysis; namely stability, reproducibility, and accuracy.
Stability: “The extent to which the results of content classification are invariant over time,” (Weber, 1990, p. 14). Stability can be determined when the same content is coded more than once by the same coder, where inconsistencies in coding constitutes unreliability. Variance might stem from ambiguities in the text, cognitive changes within the coder, or simple errors.
In order to deal with this issue, we propose to add duplicates into our dataset, which will then be coded by the analyst twice. In order to deal with possible cognitive variation, of which we believe fatigue to be the most pertinent, we will make sure that the reports are read at different fatigue levels, consistent with Krippendorff (2004) recommendations. We expect fatigue to increase with the amount of reports read in a day. For ex post analytical purposes, we will add a self-constructed variable for time of reading:
𝑀𝑒𝑎𝑠𝑢𝑟𝑒 𝑜𝑓 𝑓𝑎𝑡𝑖𝑔𝑢𝑒 = 𝑛𝑖/𝑁
Here, 𝑛𝑖 could be the 4th annual report that day, and N is the total number of annual reports read that day. We propose the Shapiro-Wilk test to be appropriate to control stability related to fatigue.
Reproducibility: “The extent to which classification produces the same result when the same text is coded by more than one coder,” (Weber, 1990, p. 17). This is the intercoder reliability, where variance is most likely to arise from cognitive differences among the coders or ambiguous coding instructions or content.
An integral part of achieving high intercoder reliability is to create an unambiguous coding scheme. In order to do this, we will follow the Weber Protocol (See the section Creating and testing a coding scheme – Weber Protocol), as proposed by Bryman and Bell (2015).
Accuracy: This refers to the extent to which the classification of text corresponds to a standard norm. Krippendorf (2004, p. 216) states that researchers seldom use accuracy in reliability assessment in research, but rather as a method for training coders, where standards are readily available.
Total assessment of reliability:
Where disagreements between coders have been found; resolving these disagreements might create bias due to differences in seniority levels between coders etc. Thus, reliability measures should be done ex ante (Krippendorf 1980, p. 132).Reliability data:
This section deals with the methods for assessing reliability. In its most canonical form, reliability data consists of the recorded data gathered by the two coders, concerning the same phenomena. In content analysis, two data-making processes are distinguishable: unitizing and coding (Krippendorff, 2004).
Unitizing:
Unitizing concerns the identification within a medium, where the sections contain information relevant to a research question. Reliability measures for unitizing are uncommon (Krippendorff, 2004, p. 220). We therefore propose proactive steps to
be taken in our scheme, and intend to include search terms as a part of the scheme, to make sure that the coders cover the same searches.
Coding
Coding concerns the interpretation of given units into data (Krippendorf, 2004, p.
220). To report the reliability of our coding scheme, we propose to use the Krippendorf’s alpha (). In its most general form, is defined by:
= 1 −𝐷𝑜 𝐷𝑒
Here, 𝐷𝑜 is the measure of disagreement between the coders, while 𝐷𝑒 is a measure of disagreement that can be expected when chance prevails. When agreement is observed to be perfect, e.g. 𝐷𝑜 = 0 = 1, which implies perfect reliability. When agreement and disagreement are matters of chance; 𝐷𝑜 = 𝐷𝑒 = 0, indicating absence of reliability. The scale could also be negative, due to sampling errors, or systematic disagreements. In terms of reliability considerations, limits are: (Krippendorff, 2004, p. 223)
1 ≥ ≥ 0 { ± 𝑠𝑎𝑚𝑝𝑙𝑖𝑛𝑔 𝑒𝑟𝑟𝑜𝑟
−𝑠𝑦𝑠𝑡𝑒𝑚𝑎𝑡𝑖𝑐 𝑑𝑖𝑠𝑎𝑔𝑟𝑒𝑒𝑚𝑒𝑛𝑡𝑠}
An alternate method for interrater reliability, is Cohen´s kappa () (J Cohen, 1960; 1968). Similar to the Krippendorff’s , Cohen´s is a chance-corrected measure (Banerjee et al., 1999). The chance-correction stems from the notion that some instances of agreement were by chance, and simple statistical methods to measure the agreement, such as the joint probability of agreement, would thus be inflated. Cohen’s is given by:
Here, 𝑝𝑜 is the relative observed agreement among coders, and 𝑝𝑒is the
hypothetical probability of chance agreement. The limits of Cohen´s is given by:
1 ≥≥ 0
It does, however, differ from Krippendorff’s when it comes to the computation of the hypothetical probability of chance. Whilst Krippendorff’s arguably offers greater flexibility in terms of number of coders and variables at different levels of measurement, the biggest drawback to its use is the complexity and
computational-heavy chance probability (Lombard et al., 2002). At this stage of our project, we therefore wish to keep the door open for both methods for measuring interrater reliability.
Validity
Two concepts of validity must be discussed. The first is validity as
correspondence between two sets of things – in our case; the correspondence between the scheme and data. Second, we will deal with validity as
generalizability of the results. These two concepts can be divided into correspondence, and generalizability (Weber, 1990).
Face validity “Face validity… consist of the correspondence between
investigators´ definitions of concepts and their definitions of the categories that measured them. A category has face validity to the extent that it appears to measure the construct it is intended to measure,” (Weber 1990, p. 19). This is the weakest form of validity, and it is likely the extent to which we will be discussing validity in our thesis. Stronger forms of validity are obtained by comparing data to some external criterion, which may be of limited supply. A brief discussion follows.
Construct validity: A measure has construct validity to the extent that it is
correlated with some other measure of the same construct. Thus, construct validity entails the generalizability of the construct across measure or methods (Weber, 1990). In the event of existing measures, any attempt to assess the construct validity of our disclosure measure will be discussed with the supervisor. Campbell
& Fiske (1959) discusses two different types of construct validity; convergent and discriminant validity. Any efforts to classify the construct validity of our measure is likely to follow this methodology.
Hypothesis validity: Hypothesis validity refers to correspondence among variables and the correspondence between these relationships and theory (Weber, 1990:20).
A measure has hypothesis validity if the relationships to other variables it behaves as it is expected to. Any discussion of this at this point is premature, but the concept can offer some guidance for our final literature review.
Sampling validity: Since our dependent variable is going to be quite time
consuming to score, any relationships that might be established between strategic disclosure and i.e. P/E ratios are not likely to be compared to other stock
exchanges that are not included in our dataset. The fact that we are covering the vast majority of Oslo Stock Exchange is likely to generate strong internal validity.
External validity is therefore going to be of less concern.
Creating and testing a coding scheme – Weber protocol
In order to create a reliable and valid scheme, we will follow the Weber Protocol (Weber, 1990), as suggested by Bryman & Bell (2015):
1. Define the recording units 2. Define the categories
3. Test coding on sample of text 4. Assess accuracy or reliability 5. Revise the coding rules
6. Return to step 3: Cycle will continue until the coders achieve sufficient reliability.
7. Code all text
8. Assess achieved reliability or accuracy.
Measurement models
To reduce the scoring from all the questions into one measure – strategic
disclosure – we may use factor analysis. This is a procedure for multivariate data that summarizes the variation of many observed or measured variables in terms of fewer underlying, or latent, variables that are called factors (Weber 1990, p58). At this stage of development, our knowledge is limited to what we have learnt from
previous studies, and, drawing upon the research that stems from the Miles and Snow (1978) Typology as discussed in the literature review, we find a continuous scale most appropriate for our research.
Challenges related to the scheme
The following section discusses methodological problems that can detract the reliability of the text classification process, and the subsequent validity of data output. We believe that there are two main challenges that need to be dealt with in order to create a reliable scheme:
Coverage:
The typical annual report from Oslo Stock Exchange is ~50-80 pages long. In order to make sure that there are no systematic differences in the way the coders assess the reports, we need to create a check list in the scheme that covers
different parts of the annual report. In that way, we believe that we can reduce the bias that might arise from cognitive dissonance.
Inference
In order to make valid inferences from the text that translate into numerical scores, we need a precise scheme. As we create our measurement variables, we need to create a generic description of scores throughout the continuum. One example could be “Does the company have an articulated strategy?”, where a low score would indicate zero, and a high score would include the pre-determined components of a clearly communicated, well-articulated strategy. In order to create this, we plan to contact linguistic and communication experts, as well as our supervisor, to discuss various configurations to best generate valid ratings.
Sample Selection
The annual report of listed firms is public, and subject to the same regulations. On the basis of this, we find listed firms to be the most appropriate for our research.
Furthermore, stock exchanges are tools for firms to reach large masses in order to raise capital. One could therefore argue that it should be in the interest of listed firms to voluntarily disclose information that might increase the value of the firm to the market, alongside what it is required to disclose.
We have initially chosen to focus on Oslo Stock Exchange, which is the primary exchange body in Norway. Although data from other stock exchanges is equally accessible, we believe the potential findings to be of a higher value to us. As it makes little difference, we are, however, be open to do our research on S&P 500 firms. Additionally, if deemed relevant and rewarding, we are open to analyzing video content related to the release of corporate annual reports. These matters will be further discussed and decided on together with our supervisor.
Panel data
Panel data contains observations of multiple entities, where each entity is observed throughout time. Panel data can be used to eliminate omitted variable bias when the omitted variable vary between firms, but not over time. One could also use fixed-effects (FE) models that eliminate omitted variable bias when variables vary over time, but is fixed across entities.
A discussion with the supervisor will hopefully advise us on the appropriate optimization of observations and points in time. One limitation regarding a pure cross-sectional analysis would be its inability to eliminate firm-specific effects, which could thus incentivize the authors to implement a method that considers multiple points in time. Additionally, we will also do initial tests of the finished scheme on a limited number of annual reports, in order to specify a suitable scope for our analysis.
Project Organization and Timeline:
The research will be conducted by two MSc in Business students at BI Norwegian Business School, as the culminating work of our degree. Our thesis supervisor is Professor Amir Sasson, who will provide guidance throughout the period of writing, as per scheduled meetings.
Creating any definite timeline of activities is difficult at this point, as it would simply be a conjecture. Thus, the proposed Gantt chart (Bryman & Bell, 2015) (See Appendix 1) should be read as a preliminary approach, and subject to ad hoc revisions. The Gantt chart does, however, illustrate the different work streams of the thesis, and helps us to divide the different parts of the thesis into manageable sub-segments.
The paper is divided into 7 main parts, that will partly overlap. The initial work will focus on constructing the ultimate research question, as well as finalizing the sample selection and research method we will use. After that, we will work on the literature review and construct the final scheme to be used, with the aim of being finalized by week 15. Much of this period will be spent on creating a
comprehensive scheme to represent strategy disclosure, as the quality of our research will to a large degree hinge on its merit. The following 10 weeks will be used to read and rate corporate annual reports, which will lead to the subsequent analysis of our findings and, ultimately, finalizing the paper by week 35.
References:
Abbott, W. F., & Monsen, R. J. (1979). On the measurement of corporate social responsibility: Self-reported disclosures as a method of measuring corporate social involvement. Academy of Management journal, 22(3), 501-515.
Abrahamson, E., & Amir, E. (1996). The information content of the president's letter to shareholders. Journal of Business Finance & Accounting, 23(8), 1157-1182.
Banerjee, M., Capozzoli, M., McSweeney, L., & Sinha, D. (1999). Beyond kappa:
A review of interrater agreement measures. Canadian journal of statistics, 27(1), 3-23.
Barron, O. E., Kile, C. O., & O'Keefe, T. B. (1999). MD&A quality as measured by the SEC and analysts' earnings forecasts. Contemporary Accounting Research, 16(1), 75-109.
Barry, D., & Elmes, M. (1997). Strategy retold: Toward a narrative view of strategic discourse. Academy of management review, 22(2), 429-452.
Bartlett, S. A., & Chandler, R. A. (1997). The corporate report and the private shareholder: Lee and Tweedie twenty years on. The British Accounting Review, 29(3), 245-261.
Bell, E., & Bryman, A. (2007). The ethics of management research: An
exploratory content analysis. British Journal of Management, 18(1), 63- 77. doi:10.1111/j.1467-8551.2006.00487.x
Benartzi, S., & Thaler, R. H. (1993). Myopic loss aversion and the equity premium puzzle. Retrieved from
Bertram, D. (2007). Likert scales. Retrieved November, 2, 2013.
Bertram, L., McQueen, M. B., Mullin, K., Blacker, D., & Tanzi, R. E. (2007).
Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database. Nature genetics, 39(1), 17-23.
Botosan, C. A. (1997). Disclosure level and the cost of equity capital. Accounting review, 323-349.
Bowman, E. H. (1976). Strategy and the weather. Sloan Management Review, 17(2), 49.
Bowman, E. H. (1978). Strategy, annual reports, and alchemy. California management review, 20(3), 64-71.
Bowman, E. H. (1984). Content analysis of annual reports for corporate strategy and risk. Interfaces, 14(1), 61-71.
Bryman, A., & Bell, E. (2015). Business research methods: Oxford University Press, USA.
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological bulletin, 56(2), 81.
Cohen, J. (1960). 0A Coefficient of Agreement for Nominal Scales, 1 Educational and Psychological MeasureM ment. XX, 1, 37M46.
Cohen, J. (1968). Weighted kappa: Nominal scale agreement provision for scaled disagreement or partial credit. Psychological bulletin, 70(4), 213.
Cooke, T. E. (1989). Disclosure in the corporate annual reports of Swedish companies. Accounting and business research, 19(74), 113-124.
Cooke, T. E. (1992). The impact of size, stock market listing and industry type on disclosure in the annual reports of Japanese listed corporations.
Accounting and business research, 22(87), 229-237.
Datta, Y. (1979). Competitive Strategy and Performance of Firms in the US TV Set Industry: 1950-1960. Paper presented at the Academy of Management Proceedings.
DeSarbo, W. S., Anthony Di Benedetto, C., Song, M., & Sinha, I. (2005).
Revisiting the Miles and Snow strategic framework: uncovering interrelationships between strategic types, capabilities, environmental uncertainty, and firm performance. Strategic Management Journal, 26(1), 47-74.
DeSarbo, W. S., & Grisaffe, D. (1998). Combinatorial optimization approaches to constrained market segmentation: An application to industrial market segmentation. Marketing Letters, 9(2), 115-134.
Dess, G. G., & Davis, P. S. (1984). Porter's (1980) generic strategies as determinants of strategic group membership and organizational performance. Academy of management Journal, 27(3), 467-488.
Diamond, D. W., & Verrecchia, R. E. (1991). Disclosure, liquidity, and the cost of capital. The journal of Finance, 46(4), 1325-1359.
Dvir, D., Segev, E., & Shenhar, A. (1993). Technology's varying impact on the success of strategic business units within the Miles and Snow typology.
Strategic Management Journal, 14(2), 155-161.
Gu, F., & Li, J. Q. (2007). The credibility of voluntary disclosure and insider stock transactions. Journal of accounting research, 45(4), 771-810.
Habermas, J. (1987). The theory of communicative action. Vol. II. Trans.
McCarthy, T. Beacon Press, Boston.
Hail, L. (2002). The impact of voluntary corporate disclosures on the ex-ante cost of capital for Swiss firms. European Accounting Review, 11(4), 741-773.
Hambrick, D. C. (1983). Some tests of the effectiveness and functional attributes of Miles and Snow's strategic types. Academy of management Journal, 26(1), 5-26.
Hatten, K. J., Schendel, D. E., & Cooper, A. C. (1978). A strategic model of the US brewing industry: 1952-1971. Academy of management Journal, 21(4), 592-610.
Healy, P. M., Hutton, A. P., & Palepu, K. G. (1999). Stock performance and intermediation changes surrounding sustained increases in disclosure.
Contemporary Accounting Research, 16(3), 485-520.
Healy, P. M., & Palepu, K. G. (2001). Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of accounting and economics, 31(1), 405-440.
Higgins, R. B., & Bannister, B. D. (1992). How corporate communication of strategy affects share price. Long Range Planning, 25(3), 27-35.
Hines, R. D. (1988). Financial accounting: in communicating reality, we construct reality. Accounting, organizations and society, 13(3), 251-261.
James, W. L., & Hatten, K. J. (1995). Further evidence on the validity of the self typing paragraph approach: Miles and Snow strategic archetypes in banking. Strategic Management Journal, 16(2), 161-168.
Kohut, G. F., & Segars, A. H. (1992). The president's letter to stockholders: An examination of corporate communication strategy. Journal of Business Communication, 29(1), 7-21.
Krippendorff, K. (2004). Content analysis: An introduction to its methodology:
Sage.
Leuz, C., & Verrecchia, R. E. (2000). The economic consequences of increased disclosure (digest summary). Journal of accounting research, 38, 91- 124No.
Lombard, M., Snyder‐Duch, J., & Bracken, C. C. (2002). Content analysis in mass communication: Assessment and reporting of intercoder reliability. Human communication research, 28(4), 587-604.
Meek, G. K., Roberts, C. B., & Gray, S. J. (1995). Factors influencing voluntary annual report disclosures by US, UK and continental European
multinational corporations. Journal of international business studies, 26(3), 555-572.
Miles, R. E., Snow, C. C., Meyer, A. D., & Coleman, H. J. (1978). Organizational strategy, structure, and process. Academy of management review, 3(3), 546-562.
Neuendorf, K. A. (2002). The content analysis guidebook: Sage.
Patten, D. M. (1992). Exposure, legitimacy, and social disclosure. Journal of Accounting and public policy, 10(4), 297-308.
Porter, M. E. (1980). Competitive strategy: Techniques for analyzing industries and competitors: Simon and Schuster.
Santema, S., Hoekert, M., Van de Rijt, J., & Van Oijen, A. (2005). Strategy disclosure in annual reports across Europe: a study on differences between five countries. European Business Review, 17(4), 352-366.
Santema, S., & Van de Rijt, J. (2001). Strategy disclosure in Dutch annual reports.
European Management Journal, 19(1), 101-108.
Shortell, S. M., & Zajac, E. J. (1990). Perceptual and archival measures of Miles and Snow's strategic types: A comprehensive assessment of reliability and validity. Academy of management Journal, 33(4), 817-832.
Stanton, P., & Stanton, J. (2002). Corporate annual reports: research perspectives used. Accounting, Auditing & Accountability Journal, 15(4), 478-500.
Strand, R. (1983). A systems paradigm of organizational adaptations to the social environment. Academy of management review, 8(1), 90-96.
Ullmann, A. A. (1985). Data in search of a theory: A critical examination of the relationships among social performance, social disclosure, and economic performance of US firms. Academy of management review, 10(3), 540- 557.
Wallace, R. O., Naser, K., & Mora, A. (1994). The relationship between the comprehensiveness of corporate annual reports and firm characteristics in Spain. Accounting and business research, 25(97), 41-53.
Weber, R. P. (1990). Basic content analysis: Sage.
Appendix
Appendix 1: Gantt chart
